I am submitting this NIDDK R03 research proposal to further the overall aims of my NIDDK K23 career development award, "Overcoming Evidence-based Barriers to Effective Diabetes Care". In the first two years of my K23 I have made rapid progress in addressing the first two aims of my proposal, which were to use existing and newly developed observational data sources to examine 1) the relationship between complexity of medical regimens and medication adherence, and 2) determinants of medication intensification (specifically the lack thereof) among patients with elevated diabetes-related risk factor levels. The insights gained from this empiric analysis of the complex process of medication management provide the basis for a series of planned pilot studies to test innovations designed to transform the current system of diabetes "usual care". The third aim of my K23 proposes to create a mechanism to link objective medication adherence data (derived from pharmacy claim refill patterns) directly to the electronic medical record used by clinicians within the Partners HealthCare System (PHS). The hypothesis of this study - to be tested in an RCT - is that these objective data will lead to more effective prescribing patterns for patients in the intervention arm. For the R03 grant entitled "Patient Medication Self-Titration for Diabetes Risk Factor Control", I am proposing to add a second pilot trial to test the hypothesis that selected patients with diabetes can effectively monitor and self-titrate oral medicines for blood pressure and glycemic control over a defined time period using a simplified, physician-approved algorithm. This pilot RCT will take advantage of the trial infrastructure being developed for the K23 study and will extend recent research by our group and others demonstrating that nurses using approved algorithms can achieve significant reductions in blood pressure and glycemic levels. We propose a two-arm RCT in which patients are randomized into either active blood pressure or active glycemia medication self-titration, with the opposite arm serving as the "usual care" corresponding control. The two pilot studies proposed for the last three years of my K23 will create the building blocks for a large R01 or R18 grant that incorporates informatics-based innovations into a new, integrated model of diabetes health care delivery that emphasizes improved communication, between-visit medication adjustments, and patient empowerment. A persistent gap exists between the effectiveness of diabetes care as currently delivered in the US and the goals of evidence-based guidelines for the ideal management of diabetes. Although patients have long been taught to self-adjust insulin regimens according to glucose self-monitoring algorithms, technological and conceptual advances now present the potential opportunity for a much larger group of patients to safely engage in oral medication self-management for hypertension and the initial stages of hyperglycemia. Increasing the patient's role in risk factor management may lead to improved diabetes control.